The Local Presence Audit Prompt, Ready to Copy

Auditing the local presence of a business is repetitive work. You check the spelling of the name, the phone, the address, and the category in a dozen places, and you write down every divergence.

It is tedious by hand and simple to describe. That is the exact shape of task that turns into a prompt.

The prompt is below, whole. Copy it and use it.

Three things to settle before you copy

The model does not reach your Maps. It does not see your profile, it does not know your rank, and it does not read your reviews. Everything it processes is what you paste into the conversation. If it states anything about your current rank, it is making that up.

It audits consistency, and not rank. What comes out of it is data divergence. Google declares literally that “there’s no way to request or pay for a better local ranking on Google”, and no prompt changes that. ¹

The output is a hypothesis to check. Every item it flags has to be verified at the source before it becomes a task.

Think of a stock clerk comparing the invoice against the shelf. He does not decide what to sell. He shows where the paper and the reality disagree. This is called a consistency audit.

The prompt

You are a local presence auditor. Your job is to compare data about the same
business across different sources and point out divergences. You have no
internet access in this task: work ONLY with what is pasted below.

BUSINESS DATA (official source, the yardstick for comparison):
Exact name:
Full address, with number and suite:
Phone:
Website:
Primary category:
Secondary categories:
Hours:

OCCURRENCES FOUND (paste below every place where the business appears,
one per block, with the origin):
[origin 1]:
[origin 2]:
[origin 3]:

RECENT REVIEWS (paste the text of 10 to 30 reviews, unedited):

TASK:

1. DIVERGENCE TABLE
   Build a table with one row per occurrence and these columns: origin, name
   as spelled, address, phone, and the field that diverges from the
   yardstick. Mark SAME when it matches.

2. SPELLING OF THE NAME
   List every distinct spelling found. Say which one is the yardstick. Flag
   any spelling that includes a slogan, a city, a service, or a word in all
   capitals, because the Google guideline asks that the name reflect only
   the real name of the business.

3. CATEGORY
   Compare the declared category against what the reviews describe the
   business doing. Point out the mismatch. Complete the sentence "this
   business IS a", and not "this business HAS".

4. CUSTOMER QUESTIONS
   From the pasted reviews ONLY, list the questions and doubts that appear,
   from most frequent to least. Quote the passage that supports each one.

5. MENTION GAPS
   From the pasted origins, say what TYPE of place already mentions the
   business and what type does NOT appear in the list.

OUTPUT RULES:
- Do not invent an occurrence that was not pasted.
- Do not state rank, ranking, or visibility. You do not have that data.
- When information is missing to answer an item, write NO DATA.
- Mark with [INFERENCE] every line that is a deduction of yours, and not a
  direct reading of what was pasted.

How to fill in the part that takes work

The occurrences block is where the audit lives, and it is the only manual work.

Search the name of the business and the phone number, separately. Every result that carries the contact data becomes a pasted block. A business directory, a trade association, a supplier site, a neighborhood page, an old catalog nobody remembers exists.

Pasting 10 occurrences already produces a useful audit. Pasting 30 produces the map.

What to do with the output

The divergence table becomes the fix list, and the fix list has an order. Start with the spelling of the name, because a divergent name breaks recognition everywhere at once. Then the phone, then the address.

The list of customer questions becomes your content plan. These are real doubts, spoken by people who already bought.

The list of mention gaps becomes your presence plan. Every missing type of place is a conversation to have.

A note on the category

The official Google guideline asks for the fewest categories: “Use as few categories as possible to describe your overall core business from the provided list”. ²

Item 3 of the prompt exists because of that. The common mistake runs the other way, and it is piling up categories in the belief that reach grows with them.

Frequently asked questions

Which model does it work in?

Any model with a context window large enough for the pasted blocks. The prompt uses no feature specific to a single tool.

Can it find the occurrences on its own?

Not as written, and that is deliberate. The prompt declares that it works only with what was pasted, because that is what keeps the output from filling up with invented mentions.

Can I ask it to estimate my rank on Maps?

You can ask, and the answer will be invention. The model does not have that data. Google also declares that no way exists to request or pay for a better local rank. ¹

What is the [INFERENCE] tag in the output?

It is the line between what the model read in the pasted material and what it deduced. Both are useful. Mixing the two is what produces a report that looks good and is wrong.

Notes and sources

1. Google, Google Business Profile Help, “Improve your local ranking on Google”, support.google.com/business/answer/7091, consulted on September 1, 2026. Literal declaration reproduced in the body.
2. Google, Google Business Profile Help, “Guidelines for representing your business on Google”, support.google.com/business/answer/3038177, consulted on September 1, 2026. Guidance on categories, literal quotation reproduced in the body. The same page determines that the name reflect the real name of the business, which supports item 2 of the prompt.